Extractor_Adaptor_Qwen3_QA_websrc_final

This model is a fine-tuned version of Qwen/Qwen3-0.6B on the web_finetune_train dataset. It achieves the following results on the evaluation set:

• Loss: 0.1143

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1.3e-05
• train_batch_size: 14
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 56
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 1.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.2497	0.0745	100	0.2823
0.0999	0.1490	200	0.1585
0.047	0.2235	300	0.1330
0.0317	0.2981	400	0.1230
0.0303	0.3726	500	0.1318
0.0321	0.4471	600	0.1310
0.028	0.5216	700	0.1143
0.0308	0.5961	800	0.1453
0.0266	0.6706	900	0.1385
0.0202	0.7452	1000	0.1351
0.028	0.8197	1100	0.1324
0.0234	0.8942	1200	0.1401
0.0268	0.9687	1300	0.1400

Framework versions

• PEFT 0.17.1
• Transformers 4.57.1
• Pytorch 2.4.1+cu124
• Datasets 4.0.0
• Tokenizers 0.22.1